display benchmark result on GH Pages

.gitignore

@@ -266,3 +266,14 @@ dmypy.json
 # Evaluation Result files
 evaluation_results/
 verification_results/
+
+# Jekyll (GitHub Pages) - docs folder
+docs/_site/
+docs/.sass-cache/
+docs/.jekyll-cache/
+docs/.jekyll-metadata
+docs/vendor/
+docs/.bundle/
+docs/tmp
+docs/Gemfile
+docs/Gemfile.lock

docs/_config.yml

@@ -0,0 +1,4 @@
+title: BC-Bench
+description: Inspired by SWE-Bench, for Business Central (AL) ecosystem.
+show_downloads: false
+theme: jekyll-theme-cayman

docs/_data/leaderboard.json

@@ -0,0 +1,52 @@
+[
+  {
+    "agent_name": "GitHub Copilot CLI",
+    "model": "claude-sonnet-4",
+    "total": 19,
+    "resolved": 7,
+    "average_duration": 0,
+    "average_prompt_tokens": 0,
+    "average_completion_tokens": 0,
+    "date": "2025-11-05"
+  },
+  {
+    "agent_name": "GitHub Copilot CLI",
+    "model": "claude-sonnet-4.5",
+    "total": 19,
+    "resolved": 9,
+    "average_duration": 0,
+    "average_prompt_tokens": 0,
+    "average_completion_tokens": 0,
+    "date": "2025-11-04"
+  },
+  {
+    "agent_name": "GitHub Copilot CLI",
+    "model": "claude-haiku-4.5",
+    "total": 19,
+    "resolved": 10,
+    "average_duration": 0,
+    "average_prompt_tokens": 0,
+    "average_completion_tokens": 0,
+    "date": "2025-11-04"
+  },
+  {
+    "agent_name": "GitHub Copilot CLI",
+    "model": "gpt-5",
+    "total": 19,
+    "resolved": 5,
+    "average_duration": 0,
+    "average_prompt_tokens": 0,
+    "average_completion_tokens": 0,
+    "date": "2025-11-04"
+  },
+  {
+    "agent_name": "mini-bc-agent",
+    "model": "azure/gpt-4.1",
+    "total": 19,
+    "resolved": 6,
+    "average_duration": 0,
+    "average_prompt_tokens": 0,
+    "average_completion_tokens": 0,
+    "date": "2025-11-05"
+  }
+]

docs/index.md

@@ -0,0 +1,38 @@
+---
+layout: default
+title: BC-Bench
+---
+
+## Evaluation Results
+
+The table below shows the performance of different AI coding agents on BC-Bench tasks. Results are sorted by accuracy (number of resolved tasks).
+
+<table>
+  <thead>
+    <tr>
+      <th>Agent (Model)</th>
+      <th>Accuracy</th>
+      <th>Avg Duration (s)</th>
+      <th>Token Consumption</th>
+      <th>Date</th>
+    </tr>
+  </thead>
+  <tbody>
+    {% assign sorted_results = site.data.leaderboard | sort: "resolved" | reverse %}
+    {% for result in sorted_results %}
+    <tr>
+      <td><strong>{{ result.agent_name }}</strong> ({{ result.model }})</td>
+      <td>{{ result.resolved }} / {{ result.total }} ({{ result.resolved | times: 100.0 | divided_by: result.total | round: 1 }}%)</td>
+      <td>{{ result.average_duration | round: 1 }}</td>
+      <td>{{ result.average_prompt_tokens | round: 0 }} prompt + {{ result.average_completion_tokens | round: 0 }} completion</td>
+      <td>{{ result.date }}</td>
+    </tr>
+    {% endfor %}
+  </tbody>
+</table>
+
+## About BC-Bench
+
+TODO
+
+For more information, visit the [BC-Bench repository](https://github.com/microsoft/BC-Bench).